USB fatal error on 2.6.35

[giuseppepezzica]

Dear all,

We are esperiencing  a USB fatal error on IMX28 BSP based on EVK28, which happens typically when We use the two USB ports (the HOST port  + the OTG port used as device) at the same time, and the throughput  is around 6 MByte/sec. It seems something related to the traffic load. When the problem happens, the host controller seems to "die". We have to reset the board in order to recover it.

Here below some indication about the  SW on the target environment :

<<<<<<

Linux version 2.6.35.3-1129-g691c08a-dirty (marco@marco-VirtualBox) (gcc version 4.4.4 (4.4.4_09.06.2010) ) #20 PREEMPT Wed Jul 29 16:09:40 CEST 2015

CPU: ARM926EJ-S [41069265] revision 5 (ARMv5TEJ), cr=00053177

>>>>>>

And here below the error description :

<<<<<<<

fsl-ehci fsl-ehci: fatal error

fsl-ehci fsl-ehci: fatal command 010038 (park)=0 ithresh=1 Async Periodic period=256 HALT

fsl-ehci fsl-ehci: fatal status cc098 Async Periodic FATAL FLR

fsl-ehci fsl-ehci: reset command 01003a (park)=0 ithresh=1 Async Periodic period=256 Reset HALT

usb 1-1: unlink qh1-0001/c7cd5700 start 0 [2/0 us]

fsl-ehci fsl-ehci: force halt; handshake c8886144 00004000 00004000 -> -110

fsl-ehci fsl-ehci: HC died; cleaning up

irq 93: nobody cared (try booting with the "irqpoll" option)

Backtrace:

[<c002ba74>] (dump_backtrace+0x0/0x10c) from [<c02b63f0>] (dump_stack+0x18/0x1c)

r7:00000104 r6:0000005d r5:00000000 r4:c039519c

[<c02b63d8>] (dump_stack+0x0/0x1c) from [<c0079404>] (__report_bad_irq+0x7c/0xac)

[<c0079388>] (__report_bad_irq+0x0/0xac) from [<c00795bc>] (note_interrupt+0x188/0x1f4)

r5:00000000 r4:c039519c

[<c0079434>] (note_interrupt+0x0/0x1f4) from [<c007a410>] (handle_level_irq+0x13c/0x180)

[<c007a2d4>] (handle_level_irq+0x0/0x180) from [<c0027044>] (asm_do_IRQ+0x44/0x94)

r6:c0389f40 r5:0000005d r4:c03a417c

[<c0027000>] (asm_do_IRQ+0x0/0x94) from [<c0027b44>] (__irq_svc+0x44/0x8c)

Exception stack(0xc0389e80 to 0xc0389ec8)

9e80: 00000000 c039433c 00000000 20000013 00000040 00000040 c03d0b80 c0388000

9ea0: c0391748 c0388000 c03d0b80 c0389f04 c0389f08 c0389ec8 c004aab8 c004a94c

9ec0: 20000013 ffffffff

r7:00000104 r6:c03d0b80 r5:f0000000 r4:ffffffff

[<c004a8f4>] (__do_softirq+0x0/0x130) from [<c004aab8>] (irq_exit+0x94/0x9c)

[<c004aa24>] (irq_exit+0x0/0x9c) from [<c0027048>] (asm_do_IRQ+0x48/0x94)

r5:0000002f r4:c03a417c

[<c0027000>] (asm_do_IRQ+0x0/0x94) from [<c0027b44>] (__irq_svc+0x44/0x8c)

Exception stack(0xc0389f40 to 0xc0389f88)

9f40: 10cc99f3 0115090d 00001000 60000013 c0388000 c03af664 c038ca30 c038ca28

9f60: 40022d60 41069265 40022d2c c0389f94 c0389f98 c0389f88 c0029690 c0034e1c

9f80: 60000013 ffffffff

r7:00000002 r6:c038ca30 r5:f0000000 r4:ffffffff

[<c0034db8>] (mx28_pm_idle+0x0/0x88) from [<c0029690>] (cpu_idle+0x9c/0xd4)

[<c00295f4>] (cpu_idle+0x0/0xd4) from [<c02b4238>] (rest_init+0xb0/0xdc)

r7:c038ca28 r6:c002501c r5:c0025020 r4:c0388000

[<c02b4188>] (rest_init+0x0/0xdc) from [<c0008a2c>] (start_kernel+0x29c/0x2fc)

r5:c0025020 r4:c03d37e0

[<c0008790>] (start_kernel+0x0/0x2fc) from [<40008034>] (0x40008034)

r5:c03af6e4 r4:00053175

handlers:

[<c01e6038>] (fsl_otg_isr+0x0/0x128)

[<c0212218>] (fsl_udc_irq+0x0/0xba8)

Disabling IRQ #93

hub 1-0:1.0: state 0 ports 1 chg 0000 evt 0000

usb 1-1: USB disconnect, address 3

>>>>>>

My question is :  someone has already experienced a problem like this ? have You any suggestions?

Many thanks,

Giuseppe

[b36401]

Did you apply all the patches?

I heard about similar issues (irq 93: nobody cared) on old builds.

Have a great day,

Victor

-----------------------------------------------------------------------------------------------------------------------

Note: If this post answers your question, please click the Correct Answer button. Thank you!

-----------------------------------------------------------------------------------------------------------------------

[giuseppepezzica · Accepted solution]

Hello Victor,

sorry for the delay answering You.

The patches We already applied in the mentioned BSP are the following :

- Patch based on L2.6.35_11_09_ER_SOURCE (REV 2011.09)

- L2.6.35 10.12 FEC0 PATCH (REV 2010.12.01)

Do You know if anything else ?

Many thanks for Your help

Giuseppe

-------------------------------------------------------------------------------------

Hello gentlemen,

We were able to  fix the problem by applying the modification indicated by Peter Chan in " Lock up of USB EHCI controller in i.MX28" , 23 Ott 2013.

Many thanks to Peter for his help.